Send buglog mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.openmoko.org/mailman/listinfo/buglog
or, via email, send a message with subject or body 'help' to
[EMAIL PROTECTED]
You can reach the person managing the list at
[EMAIL PROTECTED]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of buglog digest..."
Today's Topics:
1. Re: Openmoko Bug #2012: 2008 updated and FDOM failure to
register to vodafone Australia (Openmoko Public Trac)
2. Re: Openmoko Bug #1493: when press power button makes some
noise during suspend time. (Openmoko Public Trac)
3. Re: Openmoko Bug #1493: when press power button makes some
noise during suspend time. (Openmoko Public Trac)
4. Re: Openmoko Bug #1789: 2008.8 screen blanks and refuses to
wakeup (Openmoko Public Trac)
5. Openmoko Bug #2061: LCM (jbt6k74) doesn't go to
sleep/deep_standby modes (Openmoko Public Trac)
6. Re: Openmoko Bug #2061: LCM (jbt6k74) doesn't go to
sleep/deep_standby modes (Openmoko Public Trac)
7. Openmoko Bug #2062: xglamo doesn't implement any power
management (Openmoko Public Trac)
--- Begin Message ---
#2012: 2008 updated and FDOM failure to register to vodafone Australia
-----------------------------+----------------------------------------------
Reporter: billk | Owner: zecke
Type: defect | Status: new
Priority: normal | Milestone:
Component: Qtopia | Version: GTA02v5
Severity: major | Resolution:
Keywords: GSM Register | Blockedby:
Reproducible: always | Blocking:
-----------------------------+----------------------------------------------
Comment(by BillK):
ok, the above logreads (as the previous statement said, I need a failure
before its worth generating a logread):
a. covers a time period when the phone was suspended and an SMS (I sent
from another phone) arrived. It woke up with the display not correctly
written - no sms. Opening the messages app showed no new messages
b. covers two more messages that never arrived - probably blocked at
vodafone as no indication they arrived
c. is after a reboot and refusal to register. I didnt wait too long as it
never gets out of "user.notice root: AtChat : F : "AT-Command Interpreter
ready"" repetitions.
I restarted X and it brought up the pin dialog (tested, rxed an sms ok)
Note that a kernel error occurs just before most (all?) of the problem
spots - "user.err kernel: [ 67.570000] BUG: sleeping function called
from invalid context at kernel/mutex.c:207"
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/2012#comment:4>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#1493: when press power button makes some noise during suspend time.
-----------------------------+----------------------------------------------
Reporter: regina_kim | Owner: [EMAIL PROTECTED]
Type: defect | Status: assigned
Priority: high | Milestone: Om2008.8
Component: System Software | Version:
Severity: major | Resolution:
Keywords: | Blockedby:
Blocking: |
-----------------------------+----------------------------------------------
Comment(by ato):
Hmm. The click seems to occur actually quite a time after the kernel has
finished the resume process. I'm uncertain whether it is the amp being
enabled, or the audio codec that is generating it. In normal running
state there's no "click" on turning on and off the amp, so that leads me
to think it's the codec or something else. You can mask it (see patch
below) by delaying 2 seconds to turn the AMO on after ALSA finishes it's
resume stuff, but this probably would mean we'd miss the first part of the
ringtone on incoming call, which would be IMO even more annoying than the
"click".
Maybe it would be better to keep the amp turned off unless something is
actually playing a sound stream? That'd probably avoid this problem and
might even save some power? Please correct me if this assumption is
wrong.
{{{
diff --git a/sound/soc/s3c24xx/neo1973_gta02_wm8753.c
b/sound/soc/s3c24xx/neo1973_gta02_wm8753.c
index dbc9c18..413970b 100644
--- a/sound/soc/s3c24xx/neo1973_gta02_wm8753.c
+++ b/sound/soc/s3c24xx/neo1973_gta02_wm8753.c
@@ -58,6 +58,7 @@
#define NEO_STEREO_TO_HANDSET_SPK 9
static struct snd_soc_machine neo1973_gta02;
+static struct timer_list lm4853_wakeup_timer;
static int neo1973_gta02_hifi_hw_params(struct snd_pcm_substream
*substream,
struct snd_pcm_hw_params *params)
@@ -576,9 +577,17 @@ static struct snd_soc_dai_link neo1973_gta02_dai[] =
{
},
};
+void lm4853_wakeup(unsigned long data)
+{
+ if(lm4853_state & LM4853_AMP)
+ s3c2410_gpio_setpin(GTA02_GPIO_AMP_SHUT, 0);
+}
+
#ifdef CONFIG_PM
int neo1973_gta02_suspend(struct platform_device *pdev, pm_message_t
state)
{
+ del_singleshot_timer_sync(&lm4853_wakeup_timer);
+
s3c2410_gpio_setpin(GTA02_GPIO_AMP_SHUT, 1);
return 0;
@@ -586,8 +595,14 @@ int neo1973_gta02_suspend(struct platform_device
*pdev, pm_message_t state)
int neo1973_gta02_resume(struct platform_device *pdev)
{
- if(lm4853_state & LM4853_AMP)
- s3c2410_gpio_setpin(GTA02_GPIO_AMP_SHUT, 0);
+ if(lm4853_state & LM4853_AMP) {
+ /*
+ * Lets wait a bit before turning on the amp to prevent
the
+ * speaker from making a "pop" noise.
+ */
+ lm4853_wakeup_timer.expires = jiffies +
msecs_to_jiffies(2000);
+ add_timer(&lm4853_wakeup_timer);
+ }
return 0;
}
@@ -639,6 +654,9 @@ static int __init neo1973_gta02_init(void)
if (ret)
platform_device_put(neo1973_gta02_snd_device);
+ init_timer(&lm4853_wakeup_timer);
+ lm4853_wakeup_timer.function = lm4853_wakeup;
+
/* Initialise GPIOs used by amp */
s3c2410_gpio_cfgpin(GTA02_GPIO_HP_IN, S3C2410_GPIO_OUTPUT);
s3c2410_gpio_cfgpin(GTA02_GPIO_AMP_SHUT, S3C2410_GPIO_OUTPUT);
@@ -654,6 +672,7 @@ static int __init neo1973_gta02_init(void)
static void __exit neo1973_gta02_exit(void)
{
+ del_singleshot_timer_sync(&lm4853_wakeup_timer);
platform_device_unregister(neo1973_gta02_snd_device);
}
}}}
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/1493#comment:10>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#1493: when press power button makes some noise during suspend time.
-----------------------------+----------------------------------------------
Reporter: regina_kim | Owner: [EMAIL PROTECTED]
Type: defect | Status: assigned
Priority: high | Milestone: Om2008.8
Component: System Software | Version:
Severity: major | Resolution:
Keywords: | Blockedby:
Blocking: |
-----------------------------+----------------------------------------------
Comment(by ato):
Oops. s/AMO/amp/ if that typo wasn't clear. ;-)
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/1493#comment:11>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#1789: 2008.8 screen blanks and refuses to wakeup
----------------------------------------------------+-----------------------
Reporter: billk | Owner:
openmoko-kernel
Type: defect | Status: new
Priority: highest | Milestone: Om2008.10
Component: System Software | Version: Om2008.8
Severity: major | Resolution:
Keywords: wakeup screen display sleep suspend | Blockedby:
Reproducible: | Blocking:
----------------------------------------------------+-----------------------
Comment(by BillK):
It does, but only very occasionally, and usually when the phone has not
registered (a separate problem? - bug #2012)
Can sometimes happen if I pull usb power soon after the desktop dsplay
shows up on bootup, or if I suspend the phone (power button pressed) when
not registered, the screen wont wake up - not sure if its even suspended.
Killing apm brings it back, but must reboot immediately anyway as it
always crashes messily within a few minutes.
However, if the phone boots up properly and registers there is no problem.
Along with my other issues, its all starting to sound like race conditions
dependent on the GSM area.
BillK
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/1789#comment:10>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#2061: LCM (jbt6k74) doesn't go to sleep/deep_standby modes
-----------------------------+----------------------------------------------
Reporter: laforge | Owner: [EMAIL PROTECTED]
Type: enhancement | Status: new
Priority: normal | Milestone:
Component: System Software | Version:
Severity: blocker | Keywords:
Blockedby: | Reproducible: always
Blocking: |
-----------------------------+----------------------------------------------
The LCM itself supports three power modes, and the jbt6k74 driver exports
the switch between the power modes to userspace.
However, there seems to be no userspace power management to actually take
care of doing this.
Therefore, one suggested solution is to register with the framebuffer
notifier and switch the LCM off whenever the backlight is switched off.
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/2061>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#2061: LCM (jbt6k74) doesn't go to sleep/deep_standby modes
--------------------------------+-------------------------------------------
Reporter: laforge | Owner: [EMAIL PROTECTED]
Type: enhancement | Status: new
Priority: normal | Milestone:
Component: System Software | Version:
Severity: minor | Resolution:
Keywords: | Blockedby:
Reproducible: always | Blocking:
--------------------------------+-------------------------------------------
Changes (by laforge):
* severity: blocker => minor
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/2061#comment:1>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
--- Begin Message ---
#2062: xglamo doesn't implement any power management
-----------------------------+----------------------------------------------
Reporter: laforge | Owner: openmoko-kernel
Type: enhancement | Status: new
Priority: normal | Milestone:
Component: System Software | Version:
Severity: blocker | Keywords:
Blockedby: | Reproducible: always
Blocking: |
-----------------------------+----------------------------------------------
the glamo chip has very fine-grained power management features (such as
clock-gating to every hardware functional unit). None of those features
seem to be curently used in xglamo.
All I can find is a function to enable the clocking for a particular unit,
but no function to disable the clocking, and thus no code that ever
switches off any functional unit.
This is important when e.g. going to screen-blank situation. We can
switch off the clocks to 2D engine, the DAC block, even disable any output
of the high-frequency video signal.
And please don't tell me this is not worth optimizing because the system
will suspend to RAM soon after blanking the screen. Imagine your mp3/ogg
player running on the device. You will not suspend to ram, but still
don't want to waste power generating a video signal that nobody can see
due to backlight being blanked
--
Ticket URL: <https://docs.openmoko.org/trac/ticket/2062>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac
--- End Message ---
_______________________________________________
buglog mailing list
[email protected]
http://lists.openmoko.org/mailman/listinfo/buglog
